Publication: El problema de las condiciones iniciales en los algoritmos de estimación recursiva de modelos lineales
Loading...
Files
Official URL
Full text at PDC
Publication Date
1992
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Facultad de Ciencias Económicas y Empresariales. Decanato
Citations
Exportar
Abstract
En este artículo se proponen dos soluciones para independizar los resultados del criterio de estimación recursiva estándar de la influencia de condiciones iniciales arbitrarias. La primera solución consiste en utilizar un algoritmo recursivo corregido que descuenta el efecto de condiciones iniciales elegidas arbitrariamente sobre la estimación final de los parámetros de un modelo de regresión. La segunda solución consiste en la utilización de los filtros basados en la propagación de la matriz de información en lugar de la matriz de covarianzas. Estos algoritmos disponen, por construcción, de condiciones iniciales exactas y son robustos numéricamente.
This article proposes two solutions to allow the final recursive estimates of a regression model to be independent from arbitrary initial conditions. The first solution uses a recursive algorithm which discounts the effect of arbitrary initial conditions on the final parameter estimates. The second solution uses filters based on propagating the information matrix rather than the covariance matrix. By construction, this algorithm has exact initial conditions and is numerically robust.
This article proposes two solutions to allow the final recursive estimates of a regression model to be independent from arbitrary initial conditions. The first solution uses a recursive algorithm which discounts the effect of arbitrary initial conditions on the final parameter estimates. The second solution uses filters based on propagating the information matrix rather than the covariance matrix. By construction, this algorithm has exact initial conditions and is numerically robust.
Description
Clasificación AMS: 62J05, 62J07
UCM subjects
Unesco subjects
Keywords
Citation
Anderson, B.D.O. y J.B. Moore (1979). Optimalfiltering. Prentice-all, Inc., Englewood Cliffs, N.J.
Anderson, J.E. (1981). "Ridge estimation of house value determinants", Journal of Urban Economics, nº 9, pág. 286-297.
Bock, M.E. (1975). "Minimax estimators of the mean of a multivariate distribution", Annals of Statistics, nº 3, pág. 209-218.
Cooley, T.F., B. Rosenberg y K.D. Wall (1977). "A note on optimal smoothing for time varying coefficient problems", Annals of Economic and Social Measurement, 6, nº 4, pág. 453-456.
Garbade, K. (1977). "Two methods for examining the stability of regression coefficients", Journal of the American Statistical Association, vol. 72, nº 357, pág. 54-63.
Gruber, M.H.J. (1990). Regression estimators. A comparative study, Academic Press, INC., New York.
Hoerl, A.E. y R.W. Kennard (1970). "Ridge regression: biased estimation of nonorthogonal problems", Technometrics, vol. 12, pág. 55-67.
Jazwinsky, A.H. (1970). Stochastic Processes and Filtering Theory, Academic Press, New York.
Ljung, S. y L. Ljung (1985). "Error propagation properties of recursive least-squares adaptation algorithms", Automatica, vol. 21, pág. 157-167.
Marquardt, D.W. y R.D. Snee (1975). "Ridge regression in practice", American Statistician, nº 29, pág. 3-20.
Schmidt, P. (1976). Econometrics, Marcel Dekker, Inc., New York.
Young, P. (1984). Recursive estimation and time-series analysis. An introduction, Springer Verlag, Heidelberg.